基于三种不同分布函数的西藏东部地区极大风速估算研究
Maximum Wind Speed Estimation in Eastern Tibet Based on Three Different Distribution Functions
中文摘要：
基于西藏东部地区9个站点40～67 a的长期观测数据（最早自1953年起）与61个站点2014年后的短期观测数据，利用三种不同概率密度分布函数，估算西藏东部地区极大风速并研究其时空分布情况。结果表明：长期来看，西藏东部地区位于西风带与季风共同影响区域，主要以西北风（频率达27%）为主，99%的样本风速不超过9.0 m/s，风速超过15.0 m/s的仅占万分之1.6；对实测极大风速分别进行极值Ⅰ型、极值Ⅱ型和极值Ⅲ型分布函数拟合，发现不同时期数据拟合后的参数无显著差异；利用最小二乘法进行检验订正，可使估算的极大风速的平均绝对偏差和均方根误差大幅降低，观测数据时间序列越长整体估算效果越好；基于极值Ⅱ型函数估算的西藏东部地区极大风速分布能更好地表征该地区实际极大风速的空间分布，极大风速变化范围为13.2～30.9 m/s。
中文关键词：
西藏东部，极大风速，概率密度函数，累积概率密度，极值Ⅱ型（Frechet）分布
KeyWords：
Eastern Tibet, maximum wind speed, probability density function, cumulative density function, Frechet distribution
Abstract：
Based on the long-term observation data from 9 stations in Eastern Tibet (the earliest since 1953) and the short-term observation data after 2014 from 61 stations, three different probability density distribution functions are used to estimate the maximum wind speed in Eastern Tibet and analyze its spatiotemporal distribution. The results show that in the long term, Eastern Tibet is jointly influenced by the westerly zone and monsoon, with the northwest wind (with a frequency of 27%) as the main wind direction. 99% of the sample wind speeds are no more than 9.0 m/s, and wind speeds exceeding 15.0 m/s only account for 0.016%. Type Ⅰ, Ⅱ and Ⅲ distribution functions are used to fit the measured maximum wind speeds, and it is found that the parameters fitted for different time periods are not significantly different. Using the least squares method for verification and correction can greatly reduce the mean absolute deviation and root mean square error of the estimated maximum wind speed, and the prolonged time series of the observation data brings better overall estimation effect. The estimated distribution of the maximum wind speed in Eastern Tibet that is based on type Ⅱ function can better reflect the actual spatial distribution of the maximum wind speed in the region, with a range of 13.2 to 30.9 m/s.